ABSTRACT Autism spectrum disorder (ASD) is a highly prevalent group of neurodevelopmental disorders whose treatment efficacy is limited by poor understanding of its causal molecular mechanisms. To gain insight into the molecular basis of autism, we sequenced the genomes of families with recessively inherited autism and identified six distinct, deleterious mutations in the gene BAF53B, which segregate bi-allelically with the disorder. Since no other mutations showed this pattern of inheritance, this indicates that BAF53B loss-of-function mutations cause autism in these patients. Preliminary behavioral testing of Baf53b-/- mice further indicates a conserved role for this gene in social behaviors. Since BAF53B protein is exclusively expressed in neurons, this suggests that autism in may arise from defects in neuronal maturation or plasticity. Indeed, preliminary RNA-sequencing data from Baf53b-/- cortical mouse neurons indicate a specific requirement for Baf53b in activity-dependent transcriptional regulation. This is consistent with an early study that observed defects in activity-dependent dendritic outgrowth and aberrant mRNA expression by microarray analysis in Baf53b-/- neurons. The mechanisms through which Baf53b regulates neuronal transcription are likely to be related to its role as a core subunit of the neuronal BAF (nBAF) ATP-dependent chromatin remodeling complex. BAF complexes utilize ATP to regulate genome accessibility and nucleosome turnover, and to evict Polycomb repressive complexes. Indeed, defects in nBAF genomic targeting have been observed at subsets of genes tested in Baf53b-/- mouse neurons, leading to the hypothesis that BAF53B loss-of-function may cause autism, in part, through failures in local chromatin remodeling by the nBAF complex. In line with this, additional BAF subunits ARID1B, SMARCA2 and BCL11A have been listed as “syndromic” autism genes by SFARI, underscoring a critical, yet poorly understood function for nBAF in regulating autism-associated molecular pathways. Thus, a deeper mechanistic understanding how Baf53b and the nBAF complex regulate neuronal gene expression is warranted. The proposed study will utilize Baf53b-/- mouse neurons as a model to uncover potential mechanisms through which BAF53B loss-of-function causes autism in humans. A combination of proteomics and next generation sequencing will be employed to define Baf53b-dependent nBAF protein interactions and targeting mechanisms (Specific Aim 1), and to determine which nBAF chromatin remodeling activities are sensitive to Baf53b loss-of- function (Specific Aim 2). The conclusion of this study will reveal basic insights into gene regulation in the nervous system and importantly, will highlight potential mechanisms through which disruptions in the nBAF complex cause autism in humans. Such information is crucial to inform the development of effective methods to diagnose and treat ASD.